Scalable training of deep learning machines by incremental block training with intra-block parallel optimization and blockwise model-update filtering.

Kai Chen, Qiang Huo

Browse the full ICASSP paper archive.